【摘要】 为诊断电力机车变压器的故障类型,以提高铁路系统的安全性,提出了一种基于模糊控制的BP神经网络模型。首先应用气相色谱技术,分析变压器油中特征气体含量与故障之间的对应关系,确定电力机车变压器常见的故障类型;然后,收集各种故障类型数据,作为神经网络的训练样本数据和测试数据;最后,建立改进BP神经网络的诊断模型,实现测试数据的仿证测试。诊断结果表明,该模型在电力机车变压器故障诊断中具有良好的实用前景。更多还原

【Abstract】 To diagnose the fault types of transformer of electric locomotive for improving the safety of railway system,the BP neural network model based on fuzzy control is proposed.Firstly,by adopting gas chromatograph technology,the correspondence between content of characteristic gases in transformer oil and faults are analyzed and the common faults of transformer of electric locomotive are determined.Then,various types of fault data are collected for using as the training sample data and test data of neural network.Finally,the improved BP neural network diagnosis model is established for realizing simulation testing.The diagnosis results verify that the model possesses excellent applicable prospects in fault diagnosis for transformer of electric locomotive.更多还原